CYP3A4 inhibitors and inducers list: strong, moderate and weak, and the substrates that matter

The classification is quantitative rather than descriptive. A strong inhibitor raises the plasma exposure of a sensitive CYP3A4 substrate at least fivefold, or reduces its clearance by at least 80 percent. A moderate inhibitor produces a two to fivefold rise, and a weak inhibitor a 1.25 to twofold rise. These thresholds come from the FDA drug interaction guidance and are what product labeling refers to when it says a dose adjustment is required with strong inhibitors.

The practical value of the categories is that labeling is written in the same language. When an oncology drug label says to reduce the dose with a strong CYP3A4 inhibitor and to avoid strong inducers, you do not need a drug-by-drug interaction table; you need to know which category the other drug falls into. That single mapping resolves a large share of interaction questions.

A few combinations account for a disproportionate share of reported serious events. Simvastatin or lovastatin with a strong inhibitor such as clarithromycin or itraconazole raises the risk of rhabdomyolysis, and it is common because both prescriptions are ordinary; the statin interaction list sets out which agents avoid CYP3A4 entirely. Colchicine with clarithromycin has caused fatal toxicity, particularly in renal impairment where both clearance routes are compromised. Ergot alkaloids with strong inhibitors can cause severe vasospasm.

On the induction side the harm is quieter and often only recognized in retrospect. Rifampin started for tuberculosis or a prosthetic joint infection can drop concentrations of an immunosuppressant, a direct oral anticoagulant such as apixaban, an antiretroviral or a hormonal contraceptive below the effective range. Nothing hurts on day one. The graft rejects, the clot forms, or the pregnancy happens weeks later, and the connection back to the prescription is easy to miss.

Yes, and it is a genuine strong inhibitor at meaningful quantities rather than a curiosity. Furanocoumarins in grapefruit irreversibly inactivate intestinal CYP3A4, so the effect is concentrated on first pass metabolism of orally administered substrates and persists until the enzyme is regenerated, which takes on the order of a day or more. That is why separating the dose from the juice by a few hours does not solve it.

The substrates where it matters most are the ones with high first pass metabolism and a narrow margin: simvastatin, lovastatin, some calcium channel blockers, tacrolimus, and a number of oncology agents. Seville oranges and pomelo contain similar compounds. Most other citrus does not. It is worth asking about specifically, since patients do not consider juice a medication and will not mention it.

Inducers increase the amount of CYP3A4 enzyme the liver and gut make, so a substrate is cleared faster and its concentration falls. The strong inducers are a short and worth-memorizing list: rifampin and rifapentine, carbamazepine, phenytoin, phenobarbital and primidone, St John's wort, apalutamide, enzalutamide, mitotane, and lumacaftor combined with ivacaftor. Moderate inducers include efavirenz, etravirine, bosentan, modafinil, rifabutin, dabrafenib and nafcillin.

Two of these behave in ways that catch people out. Carbamazepine induces its own metabolism, so serum levels drift down over the first few weeks of therapy even on a fixed dose, and the dose that was right at initiation is often not the dose that holds at steady state. St John's wort is the one nobody asks about, because patients do not think of a supplement as a drug and it is rarely on the medication list. Ask about it directly in anyone on a transplant regimen, an anticoagulant, an antiretroviral or hormonal contraception.

The failure mode is silent, and that is the whole problem. When an inhibitor is added, the substrate concentration rises and the patient develops something you can see: myopathy on a statin, bleeding on a DOAC, bradycardia, confusion. When an inducer is added, the concentration falls and nothing happens that looks like a drug reaction. What you get instead is a transplant rejection, a breakthrough seizure, a rising viral load, an INR that drifts below range, or an unintended pregnancy on hormonal contraception. Every one of those reads as the disease behaving badly rather than as an interaction, so nobody goes looking for the cause in the medication list.

The timing hides it further. Enzyme inhibition is largely immediate, because the inhibitor competes for an enzyme that already exists. Induction requires the cell to synthesize new enzyme, so the effect builds over roughly one to two weeks and reaches its maximum well after the prescription was written. By the time the substrate stops working, the new drug no longer looks new.

Longer than most prescribers expect, and this is the point that changes decisions. Because induction works by increasing the amount of enzyme present, it only resolves as that excess enzyme is degraded and replaced at the normal rate. Stopping a strong inducer therefore does not restore normal metabolism within a day or two. The offset commonly takes two to four weeks, and the patient continues to clear substrates abnormally fast throughout.

The practical consequence is that timing a substrate around an inducer is not the same problem as timing it around an inhibitor. Starting a CYP3A4-dependent drug immediately after a course of rifampin ends, for example, can still produce sub-therapeutic exposure, which is why nirmatrelvir with ritonavir cannot simply be started the day a strong inducer is stopped. Plan the washout deliberately, and where the substrate has a narrow therapeutic index, monitor concentrations across the transition rather than assuming the interaction ended with the last dose.

Inhibition runs on a different clock. Competitive inhibitors wash out with the inhibitor itself, over a few half-lives. Ritonavir is the exception worth knowing, because it inactivates CYP3A4 irreversibly, so recovery depends on new enzyme synthesis and is commonly described as taking two to three days, and longer in older adults.

A great many CYP3A4 substrates are also substrates of P-glycoprotein, the efflux transporter in the gut wall and blood brain barrier, and many inhibitors hit both. Ritonavir, clarithromycin, itraconazole, verapamil and amiodarone are all dual inhibitors. When both pathways are blocked at once, exposure rises more than either mechanism alone would predict, which is why apixaban and rivaroxaban labeling singles out combined CYP3A4 and P-glycoprotein inhibitors rather than CYP3A4 inhibitors generally.

The same overlap operates in the induction direction. Rifampin induces both, which is why it reduces direct oral anticoagulant exposure so effectively. Keeping the two mechanisms in mind explains several interactions that look anomalous if you think about CYP3A4 alone.

Knowing that clarithromycin is a strong CYP3A4 inhibitor is the easy part. Noticing it at the moment you prescribe it to a patient whose home list includes simvastatin, apixaban and tacrolimus, at the end of a full clinic, is the hard part. That is a memory and attention problem, not a knowledge problem, and it is what decision-support is actually for.

Prescriber.io checks the regimen you enter for interactions including CYP3A4 inhibition and induction, states the mechanism in plain language with the source cited, and returns it alongside contraindications, allergy blockers, renal and hepatic dose considerations and guideline-based alternatives, in one card. It is decision-support for licensed US clinicians, never autonomous prescribing. You review the flag, verify against the current labeling, and sign.